1 Technological InnovationVariety in the Meaning Attributed to Invention,Innovation and Technology and the Organisationof this BookThis chapter begins with a review of definitions of technology and innovation.This is then used to develop the device of the technology complex, a device thatis exploited in the organisation of the rest of this book. This preference for thelogicaldevelopmentofanargumentfromfirstprinciplesdoesmeanthatdis-cussionoftheorganisationofthebookispostponedtohalf-waythroughthechapter.Thesecondhalfofthischapterdemonstratesthevalueofthetechnology complex as an intellectual tool by arguing for the comparative inad-equacyofthemorereadilyavailableconceptsapplicabletoinnovationandtechnological change.Invention, Innovation and TechnologyThere appear to be almost as many variant meanings for the terms invention,innovation and technology as there are authors. Many use the terms inventionandinnovationinterchangeablyorwithvaryingdegreesofprecision.Atanextreme, Wiener prefers invention to describe the whole process of bringing anovelty to the market (Wiener 1993). In contrast, Freeman prefers to restrict themeaningandincreasetheprecisionofinnovationbyonlyapplyingittothefirst commercial transaction involving a novel device (Freeman 1982: 7).Somedefinitionsareinorder.Inthisbookinventionwillberestrictedtodescribethegenerationoftheidea ofaninnovation.Innovationwilldescribesome useful changes in technology, and technology refers to the organisation ofpeopleandartefactsforsomegoal.Inthisbookthen,technologyisboththefocusofanalysisandyetisgivenaverybroadandinclusivemeaning.Thisusage is in contrast to many other authors and so deserves further explanation.The term technology is in a class of its own for variation in meaning andFigure 1.1 represents an effort to display some of this variety.The spread of definitions in Figure 1.1 has the striking quality that distinctlydifferent elements appear in many of the definitions. The titles of the works fromwhich the definitions are drawn show that the detail of the definition is linkedto the discipline, or problem of study: industrial relations, organisational behav-iour, operations management, and the problem of technology transfer. These arenot wrong definitions if one accepts the restricted focus of a subject discipline,problem or time frame and a general definition of technology should be able toincorporate such subdefinitions as special cases.chap01.qxd11/27/041:46 PMPage 12 The Management of Innovation and TechnologyCollins Dictionary1 The application of practical or mechanical sciences to industry or commerce.2 The methods, theory, and practices governing such application.3 The total knowledge and skills available to any human society.Oxford English DictionaryScienceorindustrialart;literally,thescienceoftechniquei.e.systematicknowledgeoftechnique.Technique:theinteractionofpeople/toolswithmachines/objectswhichdefinesaway of doing a particular task.Technology and Change the New Heraclitus (Schn 1967)Any tool or technique: any product or process, any physical equipment or method of doing ormaking by which human capability is extended.The Trouble with Technology (Macdonald, et al. 1985)Technology may be regarded as simply the way things are done.Technology Policy of Economic Development, IDRC, Ottawa (Vaitsos 1976)Identifies three properties of technology:1 The form in which technology is incorporated: machines/equipment/materials.2 Necessary information covering patents and conditions under which technology can beused.3 Cost of technology i.e. capital.The Management of Technology Transfer, International Journal of Technology Management(Djeflat 1987)Technologymarketedasacompleteentity:alltechnologicalcomponentstiedtogetherandtransferredasawhole:capitalgoods/materials/knowhow/qualifiedandspecialisedmanpower.The Business Enterprise in Modern Industrial Society (Child 1969)The equipment used in the work flow of a business enterprise and the interrelationship of theoperations to which the equipment is applied.Competition and Control at Work (Hill 1981)In the first place technology embraces all forms of productive technique, including hand workwhich may not involve the physical use of mechanical implements. Secondly, it embraces thephysical organisation of production, the way in which the hardware of production has been laidout in a place of work. The term therefore implies the division of labour and work organisationwhich is built into or required for efficient operation by the productive technique.The Sociology of Invention (Gilfillan 1935)An invention is essentially a complex of most diverse elements a design for a physical object,a process of working with it, the needed elements of science, if any, the constituent materials, amethod for building it, the raw materials used in working it, such as fuel, accumulated capitalsuchasfactoriesanddocks,withwhichitmustbeused,itscrewwiththeirskills,ideasandshortcomings,itsfinancialbackingandmanagement,itspurposeanduseinconjunctionwithothersidesofcivilisationanditspopularevaluation.Mostofthesepartsinturnhavetheirseparatelyvariableelements.Achangeinanyoneoftheelementsofthecomplexwillalter,stimulate, depress or quite inhibit the whole.chap01.qxd11/27/041:46 PMPage 2Use of the Technology Complex3Anexampleofhowanapparentlysimpletechnologyneverthelessincludesarange of these elements is given by the Neolithic technology of stone-napping.Technological Innovation 3Research and Technology as Economic Activities (Green and Morphet 1977)To sum up, the technology of a particular process or industry is the assemblage of all the craft,empirical and rational knowledge by which the techniques of that process or industry are under-stood and operated.Operations Management (Schroeder 1989)Thatsetofprocesses,tools,methods,proceduresandequipmentusedtoproducegoodsorservices.Figure1.1 Arangeofdefinitionsoftechnology(fromFleckandHowells2001:524,reproduced with permission of Taylor & Francis Ltd)MATERIALENERGY SOURCEARTEFACTS/HARDWARELAYOUTPROCEDURES (PROGRAMS, SOFTWARE)KNOWLEDGE/SKILLS/QUALIFIED PEOPLEWORK ORGANISATIONMANAGEMENT TECHNIQUESORGANISATIONAL STRUCTURECOST/CAPITALINDUSTRY STRUCTURE (SUPPLIERS, USERS, PROMOTERS)LOCATIONSOCIAL RELATIONSCULTUREFigure1.2 Thetechnologycomplex(fromFleckandHowells2001:525,reproducedwithpermission of Taylor & Francis Ltd)Examination of the range of definitions in Figure 1.1 suggests that a suitablegeneral and inclusive definition of technology becomes the knowledge of howto organise people and tools to achieve specific ends. This is certainly general,buthardlyuseful,becausethedifferentelementsofthesubdefinitionshavebeen lost. The technology complex1in Figure 1.2 has been suggested as a devicethat relates the general definition of technology to its subdefinitions (Fleck andHowells 2001).Theelementswithinthetechnologycomplexhavebeenorderedtorangefrom the physical and artefactual to the social and the cultural.2This capturesthe idea that there are multiple levels within society at which people organisearoundartefactstocreateworkingtechnologies.Anyoralloftheseelementscould be analysed in a working technology a technology in use. It is ratherrare that a full range of elements are considered, but it will prove worthwhile toprovide some examples of when it makes sense to extend the range of analysisover the range of the technology complex.chap01.qxd11/27/041:46 PMPage 34 The Management of Innovation and TechnologyAttheoperationslevel,skilledindividualsusebonestoolstoshapethemajor raw material flint to produce stone tool artefacts. The stone tool arte-factsthenhadawiderrangeofusespreparingskins,weaponsandwood.Theirproduction,thoughinvolvinghighlevelsofskill,appearssimpleinorganisational and material terms.However, this simplicity may be more the product of examining a simplecontexthere,routineproductionoftheartefact.Otherelementsofthetech-nologycomplexwillbeactiveandapparentifnon-routinechangesintheproduction and use of the artefact are studied.An excellent example of this is the account by the anthropologist Sharp of theeffectoftheintroductionofsteelaxe-headsintoStoneAgeAboriginalsociety(Sharp1952),whichrevealsthecomplexinterrelationshipbetweenartefacts,social structure and culture.In this patriarchal society, male heads of households buttressed their socialpositionthroughtheircontroloftheuseofstoneaxes,primarilythroughthelimitation of access to young males and women. The indiscriminate distributionofsteelaxeheadstoyoungmalesandevenwomenbywesternmissionariesdisrupted this balance of power within the tribe.SteelaxesworeawaymoreslowlythanstoneaxeheadsandthisphysicalpropertyhelpedtodisruptthetradingpatternsthatconnectednorthandsouthAboriginal tribes. The raw material for making axes existed in the south and it wasprogressively exchanged through a network of tribes for goods and materials fromthe tropical north. The annual gatherings when exchange took place had ritual,quasi-religioussignificanceaswellaseconomicexchangesignificance,butthearrival of steel axes removed the necessity to meet on the old basis and so under-minedtheculturalaspectsoftheannualritualmeetings.Inthesewayssocietyand culture were disrupted by a change in the material of an important artefact.When changes to stone axe technology were made the subject of enquiry itwasclearthatstoneaxetechnologywasnotsimpleinitssocialcontext.Withinthesocietythatgeneratedandusedthistechnologytheartefacthadcomplex significance that involved many elements of the technology complexfor its description.The technology complex warns us that what appears to be a simple technologymaybesimpleonlythroughtheconstructionofthecaseanalysis.Simplehere means describable through only a few of the elements from the technologycomplex.Modern technologies are obviously more complex at the level of the artefactand organisation and they are sustained within a more complex society. As inthe stone technology example, the study of their routine use is likely to yieldrelativelymoresimpledescriptionsthanthestudy,forexample,ofthesocialprocess of their generation or implementation. An example of the latter is theaccountbyHowellsandHineofthedesignandimplementationofEFTPOS(ElectronicFundsTransferatthePointofSale),anITnetwork,bytheBritishretail banks (Howells and Hine 1993). This found that a complex set of choicesofartefactsandsocialarrangementshadtobedefinedbythebanks.Thesechoices ranged across the full range of complex technology elements, as shownin Figure 1.3. Decisions in these categories served to define the technology ofthe IT network that was eventually implemented.chap01.qxd11/27/041:46 PMPage 4Material/artefactVariety in the designs of computer, software, communications and terminals offered imposed anexpertise management problem on the banks. Isolated artefact design decisions had unexpectedimplicationsforothernetworkcomponentdesignandcost.Thisvarietyofoffereddesignsimposed a learning process on the banks.Topology/layoutThe artefact components could be linked in different ways to represent preferences of the own-ing organisations. So transaction processing could take place in a jointly owned organisation, orthe banks IT departments. Terminals could be stand-alone or integrated into retailer equipment.Suchdecisionshadimplicationsforcompetition,ownershipandcontrolofthenetworkandchanged through the project.Procedures/softwareThereweremanywaysofbuildingsecurityintothenetwork.Onepossibilitywastoencryptelectronicmessages,butthenthereweretworivalproposalsforencryptionmethodthateachhad its own political and competitive implications.Organisational structure and location of technical expertise knowledge/skillsThe banks had to decide on how to organise the development of the network. They first experi-mented with the option to contract out responsibility for network design, then a jointly ownedhybrid organisation of technical consultants controlled by commercial bankers, then full scaledevelopment occurred through VISA and Switch network designs to which each bank made anindividual decision to affiliate.Cost/capitalThe cost of IT networks is large and the returns depend on the speed with which paper-basedsystems can be closed down. There is a theoretic role here for sophisticated financial evaluationtechniques, but the impetus for the technology derived from inter-bank competition rather thanthe immediate economics.Industry structureAsanoligopoly,thebankingindustryhadarangeofmoreorlesscooperativeorcompetitiveoptions available to it in the organisation of this large-scale project. The banks began by cooper-atingfullyintheirapproachtoEFTPOS,butcooperationbrokedownintotwofactionsthatwould develop two rival designs of network.Social/legal relationsIn 1986 the UK government passed legislation that allowed Building Societies to compete withbanks. This helped prompt the breakdown of the banks cooperative approach to network design.CulturePast experience of government led many of the banks to expect the Bank of England to regulatethe sector and to signal its approval or disapproval of bank strategies. On the EFTPOS projectsome banks continued to wait for Bank of England guidance, some interpreted Bank ofEngland statements as signalling approval or disapproval; late in the project most banks agreedthere was no longer the close supervision of earlier years.Figure 1.3 EFTPOS technology as an example of the technology complex (from Fleck andHowells 2001, reproduced with permission of Taylor & Francis Ltd)Technological Innovation 5chap01.qxd11/27/041:46 PMPage 5The Problem of the Attribution of Causality andthe Design of a StudyThese examples raise the general issue of how causation should be attributed inaccounts of technology development. It is superficially attractive to express thisproblem in a general way as the issue of how technology and society may relateto each other. In particular the question does technology shape society?, some-times described as the issue of technological determinism, has been intensivelydebatedinthehistoryandsociologyoftechnology.4Thetechnologycomplexwouldsuggestthatagoodresponsetothisquestioniswhatdoyoumeanbytechnology? For technology should be understood to necessarily include peo-ple, so that to ask how technology influences society is to ask how the peopleorganised around a particular set of artefacts influence society. Expressed thiswayitiseasiertoseethatwhetherandhowtechnologychangessocietywilldepend on the activity of both those most directly attached to the technology,forexamplethroughtheircareersorownershiprights,andthosewhowillbeaffected more indirectly, by the products or pollution that result from the devel-opment of the technology. The question how does technology affect society? isonly apparently simple, because it is short and is expressed as if there could bean answer with general validity. Any answer depends on the characteristics ofboth a specific technology and a specific society: there is no general answer.As part of a collection of writings on this topic of technological determinism(Smith and Marx 1994), the historians Thomas Misa and Bruce Bimber argue thatsubjectdisciplineandscaleofaccount(Misa)predisposeauthorstoattributecausation in the innovation process quite differently (Misa 1994; Bimber 1994).A key difference between what Misa calls micro- and macro- scale studiesisthedegreetowhichtheuncertaintyofthedecisionprocessiscaptured.Micro-level studies of managers thought and behaviour, such as that of EFTPOSdesign and implementation, capture the messiness of real processes of change;the wrong ideas, the misconceptions, the failed bids for competitive change thatformed part of the decision process. In contrast, a macro-level study of tech-nological change in retail banking over a long time span would find the seriesoftechnologiesthatwereactuallybuilttheuncertaintyandthealternativepaths of development are lost and long-run patterns of change would appear tohave their own logic, related to the cheapening cost of computer power over time.Microandmacrostudiescapturedifferentaspectsoftheprocessofinnovationand in principle they should complement one another.Tocaricaturethepossibleproblemsofsuchaccounts,unlessmicro-scaleauthorsimportvaluesfromoutsidetheirstudiestheymaybecomeunabletojudge the value of the outcomes they describe they tend towards relativism.In contrast, Misa argues that it is those whose research is conducted at the macrolevelthattendtowardsoverdeterminedinterpretationsoftheiraccounts,forexample by attributing excessive rationality and purpose to the actions of actors(Misa1994:119).HisexampleisthebusinesshistorianAlfredChandlersinterpretation of the Carnegie Steel Companys vertical integration through thepurchase of iron ore deposits. Chandler imputes a defensive motivation to verti-cal integration, a desire by Carnegie to defend his company against new entrants(Misa1994:12). Misasmicro-analysisofthedecisiontointegrateverticallyrevealsthatCarnegieandhispartnerswerelongreluctanttobuyironore6 The Management of Innovation and Technologychap01.qxd11/27/041:46 PMPage 6propertiesanddidsoonlywhenconvincedbyamiddle-rankmanager,whoseargumentswerelargelybasedontheserendipitousprofitsthathebelievedwould accrue to the company from the purchase of iron ore land land that hadbecometemporarilycheapasaresultofearliercompanyactions.AsMisacomments:Verticalintegrationprovedeconomicallyrational(lowercostsforironoreplusbarrierstoentry)forCarnegieSteel,butitwasatwo-decade-longprocesswhoserationalityappearedmostforcefullytolaterhistoriansandnot to the actors themselves. (Misa 1994: 138)With a broad concept for technology such as the technology complex, or Misasessentially similar sociotechnical networks (Misa 1994: 141), the effort to useand mix different types of account of technology development can be expectedto generate many similar problems of interpretation.Technology, its Uses and the Institutions of theMarket EconomyThe very term technology implies usefulness of some kind, for someone, mostobviousattheleveloftheartefact,anobjectbydefinitiongivenformbyhumankindandwhosecreationnecessarilyimpliessomehumanpurpose,some human use. Innovation in its widest sense is then some change in the waytechnology relates to its uses. The questions this formulation begs are of socialcontext useful how and to whom?Thisbookbeginsbyfollowingtheconventionthatmanagementstudiesshould be concerned with organisations operating within the market economyand in recent periods of time, a convention that follows the incessant demandsforthesubjecttobeimmediatelyrelevanttomanagementpractice.Thereisnothingwrongwithsuchademandinitself,butitdoesriskthatwetaketheorganisation and performance of the market economy for granted and forget thatthis is only one way of organising society and technology. One of the standardeconomics texts defines a market economy as one in which people specialise inproductive activities and meet most of their material wants through exchangesvoluntarilyagreeduponbythecontractingparties(Lipsey1989:786).Thisclarifies the nature of the economic motivation for technology development forthose directly dependent on the market economy for their financial resources to improve, or to add to the stock of market tradable goods and services. Yet itignores the institutional arrangements that are essential to a functioning marketeconomy.Institutionsandstatepolicyhavealwaysbeencentraltothesubjectofpolitical economy, and economic and business history have always found roomto include the origin and development of such institutions as industrial invest-mentbanksandtechnicaleducation.Thereisatendencyfortheimportanceofsuchinstitutionstoberediscoveredbysubjectsoncenarrowlyconceived:theeconomicsofinnovationexpandeditsscopetoincludeinstitutionsandstatepolicyinthe1990s,assymbolisedbythepublicationofacollectionofTechnological Innovation 7chap01.qxd11/27/041:46 PMPage 78 The Management of Innovation and TechnologyessaysonNationalInnovationSystems (Nelson1993);WhitleyhassimilarlysoughttoexpandthescopeoforganisationalsociologyinaseriesofeditedessaycollectionsonbusinesssystemsculminatinginhislatestbookentitledDivergent Capitalisms (Whitley 1999). As implied by these titles, institutionalendowments and state policy style vary significantly between nations and theguidingquestionformuchofthiskindofworkiswhetherthesedifferencesmatter for economy and sometimes society. The use of the word system withitsimplicationofathorough-goingmasterdesignisunfortunate,butbytheend of this book it will be obvious that if one wants to understand how and whyfirms innovate one simply cannot ignore such institutional arrangements.It is a matter of common observation that even modern societies organisedlargely as market economies contain classes of technology that are socially use-ful but that are developed outside, or uneasily alongside, the market economy forexample,themilitarytechnologiesorthetechnologiesdevelopedbythepubliclyfundedinstitutionsofscienceorganisedsothatthegenerationofunderstanding is their most immediate motivation, rather than the direct gen-eration of economic activity. Nevertheless, once technologies have been demon-stratedasviableformilitaryorscientificpurposes,inprincipletheyareavailablefordevelopmentformarketeconomicpurposes.Accordingtoitsinventor, the idea of the useful technology of the laser was motivated by a desireto create a device that would further scientific understanding (see Chapter 2).Thatisindeedhowtheinventorsoughttoapplyhistechnology,butoncea working device had been publicly demonstrated, private sector firms began todevelop the technology for economic uses.Inotherwords,eveninsocietieslargelyorganisedasmarketeconomies,while all technologies have their uses, they do not necessarily have an economicmarketandtheyarenotnecessarilyfirstgeneratedtoserveeconomicpur-poses.Theymayneverthelesscometohaveasignificanteconomicimpact.To discuss technologies and their uses in principle allows a discussion of anytechnology,whateveritssocialandculturalcontextandwhateverthehumanpurpose motivating its creation and use. To discuss technologies and their rela-tionship to the economic institution of the market implies a restriction of therange of technologies to those that generate tradable goods and services and thatserve the interests of the people and organisations participating in the market.It is primarily the second class of technologies that are of interest in this book,butasinthecaseofthelaser,non-economicinstitutionswillbeofinterestwhen they have a productive interface with the market economy.Itisonethingtobeinterestedinthetechnologiesdevelopedwithinthemarket economy and developed for economic reasons, but quite another to beuncritically committed to economic theory as an aid to understanding. Economictheory may have little to offer the study of technology. Although this book doesmake use of ideas from the subject of economics, it is also the case that certainideaswithineconomicthoughthavebeenappliedtothedetrimentoftechno-logicaldevelopment.ThisisveryapparentintheapplicationofUSantitrustandpatentlaw,discussedatlengthinChapters4and5oncompetitionandintellectual property. It should be said that the revision of policy in these caseshas also been driven by economists, but economists with particular theoreticalcommitments. The point here is that our interest in technological developmentchap01.qxd11/27/041:46 PMPage 8in the market economy can lead quite logically into an interest in the constructionandapplicationofeconomicthought,asoneofthemanyinfluencesonthedevelopment of technology.Socio-economic Perspectives on Technology and the Market EconomyThe problem of how to include economics in the study of technology while notnecessarilyacceptingtheassumptionsofeconomictheoryhasbotheredthesociologistsoftechnology.AsthetitleofacollectionofreadingseditedbyMacKenzieandWajcman,thesocial-shapingoftechnologyhasbeenpopu-larised as a term to describe some of the many ways in which technology canbe influenced by society (MacKenzie and Wajcman 1999b). Like many others,MacKenzie and Wajcman understood that neoclassical economic theory, as thedominant economic theory of the relationship between technology and econom-ics, could have little relevance when technology was changing, for then futureprices, costs and competitive scenarios matter, and these cannot be known in full,asneoclassicaleconomictheoryrequires(MacKenzieandWajcman1999a:13).Nevertheless,MacKenzieandWajcmanrecognisethateconomicsmatterstothe social shaping of technology and so they take the position that the economicshaping of technology through price and cost considerations is a form of socialshaping, and that such economic calculation depends on the organisation of thesocietyinwhichitoccurs.Thisisessentiallythesamepositionastheonedeveloped here; however, none of the examples in MacKenzie and Wajcmanstextconcernthemutualshapingbetweentechnologyandmarketeconomyinstitutions that are of interest in this book.5Ofgreaterapparentrelevanceistheimportantstreamofcontemporaryeconomicthoughtthathasattemptedtoreintroducetheanalysisofinstitu-tionstoeconomics-the-subject,theso-calledNewInstitutionalEconomics(Swedberg and Granovetter 2001: 15). However, Swedborg and Granovetter arecritical of this stream of thought as part of their development of what we mighttermacounterprogrammeforthestudyofthesociologyofeconomiclife.6AccordingtoSwedbergandGranovetter,centraltoallversionsoftheNewInstitutional Economics isthe concept of efficiency. Institutions, it is argued, tend to emerge as efficientsolutions to market failures another common feature is the assumption thatwhilemostoftheeconomycanbeanalysedperfectlywellwiththehelpofrational choice (and with total disregard for the social structure), there do existafewcaseswhereyouhavetotakenormsandinstitutionsintoaccount. Thiswayofarguingmakesforatypeofanalysisthatisonlysuperficiallysocial. Finally there is also the disturbing fact that when economists analyseinstitutions they feel no need to take into account how and why the actors,in their own view [authors emphasis], do as they do. These two limitationsinattentiontosocialstructureandtothebeliefsofindividualsmaketheNewInstitutionalEconomicsintoakindofanalysisthatcanbestbedescribed as a science of what economists think about what is happening inthe economy. (Swedberg and Granovetter 2001: 15)Technological Innovation 9chap01.qxd11/27/041:46 PMPage 910 The Management of Innovation and TechnologyThisamountstoageneralwarningthatthisschoolofeconomistshaveatendencytofindwhattheypresupposeisthere.InMisasanalysisofChandlersaccountoftherationalityofverticalintegrationintheUSsteelindustry we already have an example of the kind of bias to explanation that thismightproduce.However,oneshouldbecarefulofacceptingSwedbergandGranovetters picture of the New Institutional Economics as a clear school ofthought.ThelessonIprefertodrawisthattheinterpretationofevidenceismorefraughtthanonemighthavesupposed,andmistakesaremade.Theremedywouldbemorecareinanalysisandawarenessofthewaysthatpriortheoreticalcommitmentsmay leadinterpretationadrift.However,thereiscertainly further justification here for other ways of studying economic life thansolely through the lens of contemporary economic thought.Implications for the Structure of this BookThis book adopts such another way. The technology complex and the issue ofscale and attribution of causation will be used to organise the contents of thisbook.First,thebookseekstointroducethevariety ofreciprocalinfluencesbetween social, organisational and artefactual forms that constitutes the designand creation of technology. Second, it seeks to move between micro- and macro-levelaccountstocapturethedifferentformsoftechnologicalreality.Third,itintroduces the shaping and reshaping of relevant institutional forms as part ofthe management of innovation.Thechaptersareorganisedtoflowfromfundamentalissuesoftechnologydevelopmentandgrowthtotherelationshipbetweeninstitutionsandtheability of the private sector to develop technologies. The institutional forms ofintellectualproperty,financeandeducationthatarethesubjectofthelaterchapters tend to be more stable through time compared with specific instancesof innovation and so lean towards being taken for granted. They are neverthelesssubject to their own processes of change, in which the private sector participates,but in collaboration with other actors such as the state. They are included herebothbecausetheyinfluencetheconditionsfortechnologydevelopmentinthefirmandbecausetheyextendthecoverageofelementsofthetechnologycomplex.In sum, this book seeks to emulate the success of the social shaping texts inoffering a rich surrogate experience in the sheer variety of how social forms andhuman purpose influence technological design, but within the narrower rangeof the operation of the market economy. In response to the discussion above ofthelimitationsofanyoneschoolofthought,therewillbeapreferencefordetailed contextual analysis of change processes. This makes it more likely thatdisjunctionsbetweenempiricalmaterialandtheoreticalcommitmentscanbeteasedoutofthesources.Thisalsoimpliesthechoiceofarelativelysmallnumber of texts and cases for their richness, detail and significance, rather thanan attempt to comprehensively cover the management literature. This approachrepresentsmydefinitionofthemanagementofinnovationandtechnologicalchange. It does not pretend to offer direct help for immediate practical problems,chap01.qxd11/27/041:46 PMPage 10but if this text succeeds in its effort to make an original synthesis of material thatexplicatesanextendedunderstandingoftheprocessoftechnologicalchange,thenbecausegoodunderstandingisthebasisforgoodpractice,itwillhavevalue.Abriefreviewofsomecontrasting,alternativeapproachestoinnovationfollows. This should reinforce the value of the broad concept of technologicalchange outlined here; on this matter the problem is not one of a lack of researchvolume or of examples, but of fragmented and inadequate, often competing andoverlapping, frameworks of understanding.Some Contrasting Conceptual Approaches to theRepresentation of InnovationThe Structure of Management EducationThemostobviousroutetoprovideabroadunderstandingofinnovationinmanagement might be thought to be the MBA; so, for example, organisationalbehaviourmightbethoughttodealwiththeorganisationaldesignaspectsoftechnology, strategic management to deal with strategies for technology devel-opment, and so on.As forcefully argued by Rickards in his review of the typical MBA curricu-lum,thereisnosharedconcernwithinnovationandchange(Rickards2000).TheacademicspecialisationsthatcomprisetheMBAdidnotdeveloparoundthe object of explaining technology or technological decisions, but around theirownprofessionalagendasandmethodpreferences.Thereisnosingledefini-tion, or discussion of technology, shared between the subjects at best there arepartialandincompatibledefinitionsofthesortcollectedinFigure1.1.Takenasawhole,oneisindangerofobtainingafragmentedandcontradictoryunderstanding of technological change and the role of management.The distinct institutional history of management education compared withthe engineering subjects (in the USA and Britain) is also likely to play a part inpromotingapartialunderstandingoftechnologicalchangeandmanagement.Artefacts and their design are the preserve and principal focus of the engineeringdisciplines.7Business and management education is provided in distinct insti-tutionsandsothereisastrongtendencytodefinebusinessandmanagementsubjectsfreefromconcernwithparticularclassesofartefact;inotherwords,withoutrootsinatechnologicalcontext.Thepotentialvaluethatmaycomefrom building appropriate business and management material into a particulartechnological context is lost.Diagrammatic Models of InnovationAn example of the effort to condense and simplify understanding of innovationis the effort to build a general model of the innovation process in diagrammaticTechnological Innovation 11chap01.qxd11/27/041:46 PMPage 1112 The Management of Innovation and Technologyform. A model is a simplified representation of a more complex reality and itsobject is to make that reality more readily understood in its essential features sothat it can be used to educate, or to instruct policy.Forresthasreviewedmanyofthediagrammaticboxandlinemodelsofthe innovation process and her review makes apparent the bewildering varietywithin even this modelling format (Forrest 1991). Some of this variety derivesfrom the degree of complexity of models. Rather like the technology complex,themorecomplexmodelsweretheresultofauthorsaddingmoreelementsoftheinnovationprocesstoimprovewhatwemightcalltherepresentationalcoverageofthemodelsatthecost,ofcourse,ofthesimplicitywhichistheobject of this kind of modelling process (compare, for example, the models inFigures 1.4 and 1.5). Forrest observed that not even these complex models werecomplete,inthatitremainedpossibletothinkoffurtherelementsthatoneshould include in a model (Forrest 1991). Her minimum list of such elementsincluded:A definite pre-analysis and pre-evaluation stage, definitive feedback loops,bothinternallywithinthefirmandexternallywiththeenvironment;theindustry and life stage of the organisation within the industry; a recognitionor the environmental variables not only the marketing and technological,but the socio-cultural and political environmental variables and the internalenvironment (culture) of the firm; and the important dimensions of time andcost/resource commitment. (Forrest 1991: 450)Figure 1.4 Twisss Egg model of innovation (Twiss 1992: 25)Creativity InnovationProduct Project IdeaTHE COMPANYScientificand technologicalknowledgeKnowledgeof marketneedsProjectchampionProjectproposalEXTERNAL ENVIRONMENTINNOVATIVE INTERNAL ENVIRONMENTEvaluationsystems Analysis Strategic considerationsProjectmanagement R & D Design Production MarketingR&Ddept.Marketingdept.chap01.qxd11/27/041:46 PMPage 12Thisbeginstoresemblethelistofthetechnologycomplex,withtheaddedcomplication that the modelling form is burdened with the object of represent-ing a definite set of stages or functions and their causal relation to each other,usuallybymeansofarrows.Themorecomplexmodelsmodifyanystrongcausalconnections,forexamplebyhavingarrowspointinbothdirectionstorepresent reciprocal or iterative influence of functions, or by many of the func-tionsorstagesbeinginterlinkedbyarrowstoshowthatinnovationdoesnotnecessarily proceed in neat, ordered stages.It may not be surprising that Forrest concludes by questioning whether thispursuitofatrulygeneralinnovationmodelispossiblegiventhecomplexityandvarietyoftheinnovationprocess.Yetshedoesnotdismissoutrighttheutilityofthiskindofmodellingprocessinsteadsheurgesmanagementtocreate their own technology or industry-contingent innovation models. So thepursuit of a diagrammatic representation in a particular innovation context maybe a useful aid to clarifying thought, but Forrest warns that one should be waryof any particular model being taken too seriously and applied too rigidly as aguide to action out of the context in which it was created.Technological Innovation 13Figure 1.5 The Schmidt-Tiedemann concomitance model of innovation (Schmidt-Tiedemann1982, reproduced with permission from Research Technology Management)FUNCTIONSPHASES DIFFUSIONDECISIONS DECISION DECISIONSgoalorientedresearch(convergent)productspecificationproductdevelopmentengineeringtoolingserial productionadaptationsandTECHNICALRESEARCHCOMMERCIALsalesacquire marketshare+profitto create businessopportunitiesDevelopment ofoverhead resourcesto start innovationDeployment ofdedicatedresourcesCHAIN OF DECISIONScarry out subsequentphases of innovationPRODUCTDEFINITIONINNOVATIONCONCEPTPROTOTYPEDEMONSTRATIONKEYDECISIONMILESTONES OPTIONSMARKETINTRODUCTIONBREAKEVENmarketresearchbusinessplanscientific+technicalprogressfeasibilityexperimentaltechnicalestimatesrealizationexploitationriskstechnologydesignIINNOVATION EXPLORATIONsupport manufacturing+ salesprepare "next generation"modi-ficationsetc.post-transfersupporttechnologydesignII(transfer)predevelopmentknow-howdemandidentifiedtechnicalevaluationmarketopportunitiesidentifiedcommercialevaluationpreparation of market +sales force and servicefacilitiesto produce accordingto business planproductfunctionmarket segmentresourcespay-backrisk structureideassearchprocesses(divergent)chap01.qxd11/27/041:46 PMPage 13Technology or Manufacturing and ServicesTheverycreation ofprimary,secondaryandtertiarycategoriesofeconomicactivity was made to demonstrate the trend with which we are all now familiar:thesteadymovementofemploymentfromagriculturaltomanufacturingandnowserviceactivities8thathasoccurredindevelopedcountriesovermanydecades and centuries. Yet even an author like Bell who made these categoriesfundamentaltohisforecastsofapost-industrialsocietycommentsthatthecategory of services is a residual: that is, it is defined by not being agriculturalormanufacturingactivity(Bell1973:15).Andhereistheproblem,forifthefocusofanalysisistechnology,theartefactandassociatedsocialchange,thecategoryofmanufacturingcaptureseconomicactivitywheretheartefactisthe end result of economic activity, but services contains such diversity it iswellnighuselessasadescriptivecategory;itcontainspersonalserviceslikehairdressing,butalsosoftwareproduction,financialservices,scientificresearch, health and government services.Therealityisthatdespitethegeneralfallinmanufacturingemployment,manufacturing output continues to rise strongly in almost all industrial coun-tries, with the exception of Britain: so between 1973 and 2000, manufacturingoutputincreasedby114%intheUSAand14%inBritain(Rowthorn2001).Thewidespreadreportingofeconomicactivitythroughthecategoriesofmanufacturingandservicestendstopromotetheillusionthatthedeclineinmanufacturing employment means that the production and use of artefacts is ofdecreasingimportance.Hencethecounter-argumentmakesaperiodicreap-pearance, evident in the titles of works such as The Myth of the Post-IndustrialEconomy(CohenandZysman1994)andInPraiseofHardIndustries(Fingleton1999).Ifindustrialmeanstheabilitytoproduceartefacts,thennorich,developedcountryispost-industrialtheyaremoreindustrialthan ever.It is worthwhile to explore how the artificial division between the statisticalcategoriesofmanufacturingandserviceslargelybreaksdownifweinsistonthinking of economic activity in terms of technologies, artefacts, and their uses.Thisisperhapsmostobviouswiththeimposedclassificationofsoftwareproductionasaservice.Thisisdespitethefactthattheoutputofsoftwareproductionisamadething,anartefact,albeitwithdistinctproperties,andalsodespitethefactthattheorganisationofsoftwareproductionhasmuchincommon withtheorganisationofproductionofotherformsofartefact.Thisseemingly arbitrary act of classification allows the great recent growth in softwareproduction to add to the measured growth of service activity.Some services are specialisations in the provision of use of highly complexartefacts, and their growth and existence are clearly dependent on improve-ments in the artefact; the airline industry is a good example of such a service.Othercategoriesofservicearetightly-linked(inCohenandZysmansterm(CohenandZysman1994:33))toparticularmanufacturingindustries;theseinclude repair, maintenance and design and engineering consultancy services.Inotherwords,tightly-linkedservicesarethecharacteristicactivitiesorgan-isedaroundaspecifickindofartefactthattogethergiveatechnologyitsseparate identity.14 The Management of Innovation and Technologychap01.qxd11/27/041:46 PMPage 14Such tightly linked activities have the property that they may be organisedin-house,inwhichcasetheyareclassifiedbythestatisticiansasmanufactur-ing,butiftheyareorganisedseparatelytotheproductionofartefactsintodays jargon, if they are outsourced they are classified as services. One canimaginethatiftherewereanetincreaseinoutsourcingintheeconomy,itwouldincreasethemeasuredstatisticaltrendtowardsmoreserviceactivityyetforsuchservices,theorganisationalshellfromwhich,orwithinwhich,theyareprovidedislessimportantthanthattheyexistatallandthattheycontinue to relate to a specific form of artefact.Indeed,inordertomakesenseofrecentorganisationaltrendsincomplexproduct industries, some writers have dropped the terms manufacturing andservicesinfavourofthebusinessofsystemsintegration(forexample,Prencipeetal.2003).ItiswellknownthatcompaniessuchastheUScarmanufacturers and General Electric have moved downstream into high value-addedservicessuchastheprovisionoffinancetosupplementtheirproductincomes,butinthe1990sGeneralElectriccontinuedtoexpandtheservicesofferedbyitsfinancialservicesdivision,GECapital,sothatitcouldofferintegrated solutions to its customers needs, including products, finance andmaintenance(Davies2003:338).By2002GECapitalgenerated49%ofthefirms total revenue (Davies 2003: 338).Ericsson offers a more radical case, where much of its product manufactur-inghasbeenoutsourcedtoafavouredmanufacturingcontractor,Flextronics,while service offerings and business consulting activities have been combinedinto a new division, Ericsson Global Services, created in 2000 with the object ofthe provision of high-value integrated systems and services to all mobile phoneoperators (Davies 2003: 352). As some of Davies cases show, while such firmsasthesearemovingfrommanufacturingintointegratedsystemsprovision,other firms have moved from being pure service providers into integrated sys-tems provision: this is the case with WS Atkins, the former project managementand technical consultancy provider (Davies 2003: 352).What this reveals is that the appropriate bundle of activities for containmentwithinthefirmisverymuchinfluxandverymuchafocusofmanagementattention. Our focus of enquiry is management activity in a technological con-text and it will prove hardly possible not to refer to the specific services thatare tightly linked to artefacts. The custom of categorising economic activity bytheendproductofthefirmsothatfirmsareeither manufacturingor servicefirms obscures more than it reveals.Reading from the Artefact Alone Common Sense and Some LimitationsIt is an interesting exercise to consider what can be learnt from artefacts alone.An instructive and contrasting pair of thought experiments consists of trying toisolate significant innovation that on the one hand has no artefact componentand on the other hand concerns the artefact alone.So, for example, if the effort is made to think of a pure marketing innova-tion,itispossibletoimaginechangesinmarketingtechnique,butonesoonTechnological Innovation 15chap01.qxd11/27/041:46 PMPage 15realisesthatasmarketingisaservicetothemajoractivityofthefirm,wherethatactivityinvolveschangingartefacts,marketingactivityislikelytobeadapted to the nature of such changes. So the search for pure marketing inno-vationshouldseekexampleswherethereisaminimumofartefactchangeinevidence. It is difficult to beat Fosters nice, laconic account of what innovationamounts to in cola drinks:We all remember the familiar 6.5 Coke bottle and its Pepsi equivalent. Andwecanalsorememberwhenthe6.5ouncebottlewasjoinedongroceryshelves and vending machines by the 12 ounce can. This advance and somepeopleconsidereditonewassosuccessfulthatitwasfollowedwiththe16 ounce glass bottle, and then the 26 ounce bottle, the 28 ounce bottle andfinallythe32ouncebottle.Fromtherethecompetitiongotdirty.Firstonecompanywenttothe48ounce,thentheother.Thenbothfollowedquicklywiththe64ouncebottle.Wherewouldallthisexcitingcompetitionlead?Tothemetricsystemwhenthe64ouncebottlewasreplacedbythe2litrebottlenowmadeofplastic,andthenbythe3litrebottle. Whatnon-advance can we expect now? Five litre bottles with wheels? (Foster 1986: 2545)Such a business is driven by advertising and marketing as the means of manipu-lating consumer perceptions of symbolic and minor changes to the presentationof the product. These activities are interesting in their own right, but the scopeformaterialchangethatmightadvantagetheconsumerisobviouslylargelyabsent.Indeed,theexamplesharpensanotherquestion:whatcanadvantagemean in the absence of some element of artefact change?The complement of the imagine innovation without the artefact exercise isto consider the meaning of a change in the artefact alone. But this is an absurd-ity because by definition change in an artefact is a result of human intention andactivity. However, these activities are often complex with an elaborate divisionof labour and expertise; so this question can be changed to become a search forthoseindividualsthatexperiencechangesinartefactswithaminimumofaccess to the human processes upon which these rely.Common sense suggests that in our society it is in our role as consumers thatwecomeclosesttoexperiencingartefactualchangewithoutthealliedknowl-edge of production and distribution. More than this, it is because of our indi-vidualsocialandcognitivelimitationsthatwehavenecessarilyrestrictedknowledgeofthediversityandcomplexityofdesignandproductionofthesesame artefacts. There is a fundamental asymmetry between our degree of accessto consumer and to producer technology.Andyet,asconsumers,wehavetheknowledgeofuseofmanyartefacts.Wearefamiliarwiththeproblemsofacquiringthisknowledgeofusetheseartefacts must be set up, used and maintained and they all fail in various ways.Whentheyfailwetendtomakejudgementsaboutthehumandesignprocessthat produced the artefact and a common struggle that occurs is over the attri-butionoffaulteithertoartefactdesignortoourmannerofitsuse.Inotherwords, we are all to some degree experienced in the interpretation of artefactsasindicatorsofthedesignandproductionprocess,despite ourintrinsicallylimited ability to grasp the detail of all these production processes.16 The Management of Innovation and Technologychap01.qxd11/27/041:46 PMPage 16All this is basic and perhaps obvious, but it is worth restating here becauseit helps avoid the mystification of technology. The imaginative basis for under-standingthedesignprocessinitsvariouscontextsiswidespread,yetusuallylimited by having access only to artefact forms. Of course experts in methods ofproduction and design should make the better interpreters of artefacts what isreverseengineeringbutahighlyandactivelydevelopedabilitytoreadfromartefact design the nature of the production and design process that created thatartefact? And outside engineering we find the discipline of archaeology to be thestudy of human history through the science of reading defunct artefacts (andother material remains).Limits on the Interpretation of the Artefact Skeuomorphs and SpandrelsFromsuchdisciplinesintensiverelationshipwithartefactformandmeaningwecanderiveusefulandgenerallimitingconceptsonwhatisprobablyourdefault assumption about any artefact form that every element of design thatwe observe in an artefact is there by the designers intention and for utilitarianreasons.A skeuomorph is an element of design that has lost its original function buthas nevertheless been retained in its artefact. The classic example in architec-tureisthesquareatthetopofthemuchimitatedDoriccolumn(Adam1989)andreproducedinFigure1.6.EarlyGreekarchitecturewasconstructedfromwoodandthissquarehadastress-distributingfunctionforwoodencolumns.Theretentionofthesquarewhenmarbleandstonematerialswereusedinsteadofwoodwasforreasonsofstyle.AnotherexampleistheretentioninTechnological Innovation 17Figure 1.6 The square at the top of the Doric column is the skeuomorphchap01.qxd11/27/041:46 PMPage 1718 The Management of Innovation and Technologymoderndust-pressedclayplatesofthediscshapeoncenecessarilyobtainedfrom wheel-spun pottery production. Many other shapes, such as squares, are inprinciple possible with the dust-pressing manufacturing process (Adam 1989).Adam suggests that the freezing of form in the plate may be partly aesthetic andpartly prevalent usage; the spread of dishwashers designed around the standardform in crockery begins to limit the latitude in crockery design.Aquitedistinctclassoflimitationsonourabilitytoworkoutthedesignprocessfromitsresultantartefactisthatwhatmayappeartotheartefactobserver as an element of design may be the byproduct of other design choices.Thedifficultyofdistinguishingbetweenbyproductandintentionaldesignelements,withintheartefactaloneandwithoutproductionknowledge,isdemonstratedinacontroversyoverthesignificanceofthespandrelasabyproduct of design in architecture.AmuchdiscussedpaperinevolutionarybiologybyGouldandLewontinused the architectural spandrel to argue that elements of biological form mightalsobebyproductsofotherdesignelements,ratherthanadaptationsundernatural selection, as conventionally assumed (Gould and Lewontin 1979). Theirpaperhasbeennotoriouslymisusedtounderminethepublicperceptionofevolution,butthatcontroversydoesnotconcernushere.9GouldandLewontins use of the spandrel in architectural form as a straightforward exampledoes concern us and they comment that such architectural constraints abound,and we find them easy to understand because we do not impose our biologicalbiases upon them (Gould and Lewontin 1979: 581). On the contrary, they arenot so easy to understand and for the same reason that the attribution of func-tionmaybecomplicatedinbiology:casualobservationoftheartefactorthebiologicalformalone,withoutsufficientunderstandingoftheprocessofcon-struction or growth, is not sufficient to determine the function of all elementsof form.Gould and Lewontins example of spandrels is the Basilica of San Marco inVenice,whichhasamaindomesupportedbyfourarches.Firstitmustbepointed out that the correct name for the roughly triangular segment of a spherethattaperstopointsatthefeetofitssupportingarchesisapendentive,notaspandrel (Dennett 1996: 3). The difference between a spandrel and a pendentiveis evident when Figures 1.7 and 1.8 are compared.Nowthehistorianofarchitecturaltechnology,RobertMark,hasusedargumentsdrawnfromhisunderstandingofthen-contemporaryconstructionknowledgetojudgethatthependentivesofSanMarcodo,afterall,haveafunction.In the earlier building (537 AD) of Justinians Hagia Sophia in Constantinople(Figure1.8),Markfoundthatpendentiveshadthefunctionofstabilisingthedome,whichwouldotherwisebeindangerofexplodingoutwardsundertensile stressderivedfromitsownweight10(Mark1996;MarkandBillington1989: 308). The pendentives performed this function when their outside surfacewasloadedwithmaterialtoprovideastabilisingweighttheseweightsarevisible on the outside of the dome in Figure 1.8. Mark then argues that becauseconstruction knowledge in the prescientific era was developed by observationof the behaviour of existing buildings, the builders of the Basilica of San Marcointhelateeleventhcenturywouldhavebeenextremelyunlikelytouseanychap01.qxd11/27/041:46 PMPage 18other construction method than pendentives to stabilise its large, 30 m diameterdome(Mark1996).Inotherwords,intheBasilicaofSanMarco,pendentiveswere not a necessary architectural byproduct (Gould and Lewontin 1979: 581)that resulted from the design choice of arches to support a dome, but they werea necessary structural feature of the proven system of support for large domedbuildings: arches with loaded pendentives were the necessarily combined ele-ments that together achieved the purpose of stabilising a dome. This leaves theGouldandLewontindiscussionofspandrelsnotonlywronginitsuseofTechnological Innovation 19Figure 1.7 The roughly triangular area created when a viaduct is supported by arches is thespandrelFigure1.8 DiagramoftheconstructionoftheHagiaSophia.Thependentivesarethecurving, three-dimensional and tapering triangular areas between the arches and the rim ofthe dome. The tensile stress within the dome is stabilised by blocks of masonry that rest onthe pendativesWeight of masonry stabilisesthe domeLoad bearing pendentivechap01.qxd11/27/041:46 PMPage 19terms, but wrong in its conclusions on the significance of byproduct design inboth architecture and evolutionary biology.11Thesignificanceofthisisthateveninarchitecture,whoseproductisthatmost visible of artefact forms, the building, mere inspection of the artefact is nota reliable means of working out the constraints on the design and constructionprocess. In this example, the interpretation of architectural form was assumedtobesimple;inreality,theproblemofinterpretationofarchitecturalformproved as complex as some of the examples that could have been provided fromevolutionary biology.In conclusion, this pursuit of the feasibility of interpretations of the artefacttendstoreinforcetheconclusionthatifinnovationismostinterestingwhenthere is an element of artefactual change, the understanding of such change isnecessarilyaccompaniedbyanunderstandingofchangeinotherelementsofthe technology complex. The attainment of that understanding is often obscureand difficult to access, but in this book, in the search for understanding of pat-terns in technological change, we will be repeatedly drawn into investigationsof the detail of the relevant technology, understood in its broad sense throughthe technology complex.ConclusionIn this review of alternative conceptions of innovation, the problem revealed isnotnecessarilythatinnovationisinaccessibleandneglected,butthatmanyreadily available conceptions and straightforward daily observation, give at bestonly a limited view of the innovation process. This is perhaps not surprising,butitreinforcestheconclusiondrawninthefirstpartofthischapter:thatifgeneralunderstandingistheobject,thenabroadconceptionoftechnologicalchange is the appropriate one. It follows, then, that the rest of this book shouldbe dedicated to an analysis of change in the elements of the broad definition oftechnology.Notes1Fleck developed the idea as a result of many years of teaching the management of technologyat the University of Edinburgh. In 2001 he and I published a paper describing the technology com-plex and giving examples of its utility. Much of this section and the following section on the usesof the technology complex are from that paper (Fleck and Howells 2001).2Thelistcouldbeexpandedthroughtheadditionofmoresubdefinitionsoftechnology,forFleck and Howells did not attempt a complete review of disciplines with active and distinct usesof the term technology. Such additions and the pursuit of an ideal complete review would notchangetheessentialpatternrevealedhere:thatagiventechnologyincludeselementsthatrangefrom the artefactual through various social forms.3This section is almost entirely derived from Fleck and Howells (2001: 5256).4For example see Smith and Marx 1994; MacKenzie and Wajcman 1985.5AbranchofthesociologyoftechnologybasedonthewritingofCallonandLatourtermedActorNetworkTheory(ANT)seemstobeincreasinglypopularwithprospectivePhDstudentsasifitofferedacoherenttheoryoftechnology.Oneoftheirvirtuesisthattheyareinfavour of more thick description of technological development, but they actively develop a jargon20 The Management of Innovation and Technologychap01.qxd11/27/041:46 PMPage 20alloftheirownthatmakestheirwritingoftenimpenetrable(Callon1991;Latour1991).Oneexample at the heart of their work is to allow non-humans by which they almost always meanartefacts to be called actors within certain stories. Callon gives the example of a story that mightbetoldaboutradioactivefalloutfromChernobyl,wherehesuggeststhattheChernobylreactorbecomesanactorinthestory(Callon1991:142).Thisextensionoftheusageofactorthenbecomesperfectlygeneralhenceactor-networksareaboutthechainsofrelationshipsbetweenhumans and non-humans or rather humans and artefacts. The result of this generalisation of anextension in usage of the word actor is that a decent English word that once usefully distinguishedthe human ability to act from the artefacts inability to act in like manner no longer can perform thistask. This example of actor is of course only the first of many other imprecise inventions and shiftsin meaning given to a run of words such as actant, inscription, intermediary. I remain scepticalof the value of this jargon. All these authors want to do is tell stories about technology and to tellthe rest of us to tell stories about technology. If historians, economists and other sociologists can tellsuch stories without the burden of the ANT jargon, it is probably unnecessary. One should also keepinmindAndreskisbookSocialSciencesasSorcery,payingspecialattentiontothechapterThe Smoke Screen of Jargon (Andreski 1974). There is a growing critical literature on ANT forexample, Winner criticises ANT for its disdain for the many other writers and disciplines that havecontributed to the serious study of technology in society and for its relativism its abandonmentofageneralpositiononthevalueofthesocialandtechnologicalchangesthatitseekstostudy(Winner, 1993).6FromthetitletoGranovetterandSwedbergseditedcollectionofessays(GranovetterandSwedberg 2001).7In this institutional arrangement, engineers with MBAs would approximate to the few withexposure across the range of relevant technology-shaping subjects. See Chapters 7 and 8 for a moredetailed discussion.8Bell credits Colin Clark with creating the terms in his book Conditions of Economic Progress(Bell 1973: 14) and uses the criterion that a majority of a countrys workforce be employed in theservice sector as his first and most simple definition of that unfortunate term, the post-industrialsociety.9As Dennett has argued, most evolutionary biologists would welcome the point that biologicaldesign is not exclusively a product of adaptation to environment. The controversy arises because,in Dennetts words, Gould has persistently misrepresented the import of the Gould/Lewontin paperoutsidebiology,andmanyhavebeentakenin(Dennett1996:3).Themisrepresentationistheimplication that one need not search for an adaptive explanation for complex biological attributessuch as language (Dennett 1996). The significance of spandrels in biology appears to be the same asin artefacts elements of biological design may result as byproducts of other processes, but theremust be other processes there must be conscious design in artefacts and there must be adaptationunderselectioninbiology.Spandrelsaremerelyevidenceofconstraintstotheseprocessesandserve only to warn us not to make over-hasty interpretations of the design process from the designed(or evolved) thing.10Mark and Billington used computer modelling techniques to establish the stress patterns inthe domes of the Roman Pantheon and the Hagia Sophia. These stresses allowed them to infer thefunction of the loaded pendentives (Mark and Billington 1989).11ThishasbeenpointedoutbyDennettinhisarticleonTheScopeofNaturalSelection(Dennett 1996: 2).Technological Innovation 21chap01.qxd11/27/041:46 PMPage 21